当前位置:文档之家› 同济大学数值计算课件

同济大学数值计算课件


A [1 2 3]
Column vector:
A [1 2 3]

Column vector:
A [1; 2;3]
A [1 2 3]'

Matrix:
1 2 3 A 4 5 6 7 8 9
A [1 2 3 456 7 8 9]
二、The future of scientific computing

Sep. 1996, America DOE proposed “ASCI” plan(加速战略计算创新) The object of ASCI: In computer platform with 1014 Flop/s ( 100万亿次/秒), 3D complete physics processing, complete scalable, high precision scientific simulation computing will be done.
四、How to learn?

The teaching style: idea + practice Point policy : exam + three projects 70 points + 30 points
五、Relation between Matlab and Numerical method
Climate Prediction
The three kernel problems in scientific computing:

suitable mathematical modeling high efficiency numerical method high performance computer platform
一、The position of scientific computing
1.


In scientific research The third eye of the modern scientific research. The three methods of the modern scientific research: Theory Experiment Scientific computing

No.1 in Shanghai(94) Magic cubic( 魔方) 0.18 Pflops/s
high performance computer platform --- Histoty


Top1 (Jun.18, 1993)-------60GFlops Top1 (Jun.18, 2012)-------16,324,751GFlops 19 years--------2.7e5
二、The future of scientific computing

Origin: scientific computing method will take place experiment method in scientific research.
二、The future of scientific computing
2. numerical computing of certain quantities
Differentiation (4) 微分 Integration (5) 积分 Ordinary Differential Equations (8) 常微分方程 Systems of Linear Equations (6) 线性方程组 Locating roots of equations (3)方程求根



Design: using CAD software to design to outer curve of car or plane. The two preliminary factor: beautiful smaller drag force Computing: using computational air dynamics, obtaining the drag force of car or plane. Judgement: if the computed drag force > the required drag force, then repeat step (1),(2); else exit from the iteration loop, stop.
2. In engineering and industrial application ----becoming wider.
The conclusion comes from IT TOP 500.
二、The future of scientific computing

is beautiful but the path is full of difficulties

I can compute traces of planets, but I can‟t imagine madness of human being. ---------------Newton
三、outline of the course

1.
the three preliminary problems of numerical computing: the problem of representing functions within a computer Interpolation (4) 插值 Approximation by Spline Functions (7) 样条 近似 Smoothing of Data and the method of least squares (10) 数据光滑和最小二乘方法

Relation between Calculator and the basic arithmetic operations addition, subtraction, multiplication, division.
Ax b
x A\b
Chapter 1 Basics of Matlab

Let us image the style of scientific research after “ASCI” plan is finished: (As an example, car‟s or plane‟s outer curve design)
An iterative process with three steps
3.numerical computing of random quantities Monte Carlo methods and simulation
四、How to learn?



Mathematics is classified into three categories: Baby mathematics + Cemetery mathematics + Church mathematics Baby mathematics emphasizes on new ideas Cemetery mathematics emphasizes on proof Church mathematics emphasizes on comment
3.Arrays with Special Structure
A 1:10
A 2: 0.5: 4
4. Special Matrices
A eye(3)
A ones(3, 4)
A zeros(2, 4)
5. Matrix Operations
C A B
Importance of scientific computing:


Foreign university, since 1970. Chinese university, since 1990. Tongji University, since 1999.
一、The position of scientific computing
M-file Example
%m-file to compute Fibonacci numbers f=[1 1]; i=1; While f(i)+f(i+1)<1000 f(i+2)=f(i)+f(i+1); i=i+1; end
8. Plotting of Graphs
x=0:0.2:3; y=x.^2&#hod and computer algorithm
陈素琴 同济大学数学系
References:
1. Applied Numerical Methods for Engineering and Scientists, Singiresu S.Rao, Prentice Hall, Upper Saddle River, NJ07458, 2002. 2. Numerical Analysis, David Kincaid, Ward Cheney, Brooks/Cole Publishing Company, 1996.
1.
Matlab: Matrix Laboratory, 1970s Variables (变量) Variables name: start with a letter and can have a length of up to 31 characters. Type: matrices.
2. Arrays and Matrices (数组和矩阵) Row vector:
相关主题